Class DiscreteEnumFromDirichletOp<TEnum>

Provides outgoing messages for DiscreteEnum<TEnum>(Vector), given random arguments to the function.

Inheritance

Object

DiscreteEnumFromDirichletOp<TEnum>

Inherited Members

Object.Equals(Object)

Object.Equals(Object, Object)

Object.GetHashCode()

Object.GetType()

Object.MemberwiseClone()

Object.ReferenceEquals(Object, Object)

Object.ToString()

Namespace: Microsoft.ML.Probabilistic.Factors

Assembly: Microsoft.ML.Probabilistic.dll

Syntax

[FactorMethod(typeof(EnumSupport), "DiscreteEnum<>", new Type[]{})]
[Quality(QualityBand.Stable)]
public static class DiscreteEnumFromDirichletOp<TEnum>

Type Parameters

Name	Description
TEnum	The type of the enumeration.

Remarks

This class provides operators which have Enum arguments.
The rest are provided by DiscreteFromDirichletOp.

Methods

AverageLogFactor(TEnum, Dirichlet)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(TEnum sample, Dirichlet probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Dirichlet	probs	Incoming message from Probs.

Returns

Type	Description
Double	Average of the factor's log-value across the given argument distributions.

Remarks

The formula for the result is sum_(Sample,Probs) p(Sample,Probs) log(factor(Sample,Probs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

AverageLogFactor(TEnum, Vector)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(TEnum sample, Vector probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Vector	probs	Constant value for Probs.

Returns

Type	Description
Double	Average of the factor's log-value across the given argument distributions.

Remarks

The formula for the result is sum_(Sample) p(Sample) log(factor(Sample,Probs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

LogAverageFactor(TEnum, Dirichlet)

Evidence message for EP.

Declaration

public static double LogAverageFactor(TEnum sample, Dirichlet probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Dirichlet	probs	Incoming message from Probs.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(Sample,Probs) p(Sample,Probs) factor(Sample,Probs)).

LogAverageFactor(TEnum, Vector)

Evidence message for EP.

Declaration

public static double LogAverageFactor(TEnum sample, Vector probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Vector	probs	Constant value for Probs.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(Sample) p(Sample) factor(Sample,Probs)).

LogEvidenceRatio(TEnum, Dirichlet)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(TEnum sample, Dirichlet probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Dirichlet	probs	Incoming message from Probs.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(sum_(Sample,Probs) p(Sample,Probs) factor(Sample,Probs) / sum_Sample p(Sample) messageTo(Sample)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio(TEnum, Vector)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(TEnum sample, Vector probs)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Vector	probs	Constant value for Probs.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(sum_(Sample) p(Sample) factor(Sample,Probs) / sum_Sample p(Sample) messageTo(Sample)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

ProbsAverageConditional(TEnum, Dirichlet)

EP message to Probs.

Declaration

public static Dirichlet ProbsAverageConditional(TEnum sample, Dirichlet result)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Dirichlet	result	Modified to contain the outgoing message.

Returns

Type	Description
Dirichlet	result

Remarks

The outgoing message is a distribution matching the moments of Probs as the random arguments are varied. The formula is proj[p(Probs) sum_(Sample) p(Sample) factor(Sample,Probs)]/p(Probs).

ProbsAverageLogarithm(TEnum, Dirichlet)

VMP message to Probs.

Declaration

public static Dirichlet ProbsAverageLogarithm(TEnum sample, Dirichlet result)

Parameters

Type	Name	Description
TEnum	sample	Incoming message from Sample.
Dirichlet	result	Modified to contain the outgoing message.

Returns

Type	Description
Dirichlet	result

Remarks

The outgoing message is the exponential of the average log-factor value, where the average is over all arguments except Probs. The formula is exp(sum_(Sample) p(Sample) log(factor(Sample,Probs))).